On June 30 2020 the city of Gent migrated its Open Data portal to a new platform, so this hack is not needed anymore (and it won't work anyway). This repository is archived and read-only now.

The Open Data portal of the City of Ghent has a collection of datasets can can be downloaded in different formats, like XML, KML, JSON or GeoJSON. The problem with the GeoJSON responses is that they normally only contain MultiPolygon or MultiPoint collections of coordinates and very little extra information (it seems the City of Ghent is not using the latest version of The DataTank, that fixes this).

This simple application reads the associated KML files to generate more complete GeoJSON responses.

You can access the data at https://geogent.herokuapp.com/. To make a request you just need to replace the City of Ghent DataTank domain datatank.stad.gent with geogent.herokuapp.com. For example, the following URLs:

https://datatank.stad.gent/4/cultuursportvrijetijd/speelterreinen.geojson
https://datatank.stad.gent/4/mobiliteit/parkinglocaties.geojson

become:

https://geogent.herokuapp.com/4/cultuursportvrijetijd/speelterreinen.geojson
https://geogent.herokuapp.com/4/mobiliteit/parkinglocaties.geojson

geogent uses Flask. Feel free to contribute to the project.

The application uses Pipenv to manage Python packages:

$ pipenv install --dev
$ pipenv shell

Update dependencies (and manually update requirements.txt):

$ pipenv update --dev && pipenv lock -r

$ python runserver.py

$ black geogent

Use Heroku Containers to deploy the app:

$ heroku login
$ heroku stack:set container
$ git push heroku master